jevmoji
Web app that has Jev score emoji categories and candidates 0 to 3 for relevance to a short text and returns many matches.
Overview
Jevmoji is a small Cloudflare Worker web app: type up to 40 characters and get a list of related emojis scored by Jev, many matches rather than a single pick. It uses a generated Unicode emoji catalog with emojibase keywords and asks Jev to score emoji groups for relevance on a 0 to 3 scale, paging through large catalog groups.
A code-built shortlist of 'hits', made by matching the text against emoji names and keywords, is scored alongside the categories. The output view includes a score trace showing which groups were scored. The TypeSafe key stays server-side in the Worker. It is a demo of using Score questions for multi-label retrieval rather than a single choice, with a live instance and unit tests.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Score | How related is this emoji (or group) to the input text, from 0 to 3? | 0 · 1 · 2 · 3 |
How it works
- User types up to 40 characters
- Worker builds candidate emoji groups and a keyword shortlist
- Jev scores candidates 0 to 3 for relevance
- Higher-scoring emojis are returned with a score trace
Strengths
- Returns several relevant emojis, not one
- Key never reaches the browser
- Catalog regenerated from Unicode data by script
Caveats
- Small demo app
- Every query is a paid Jev call
- Input limited to 40 characters
Getting started
npm install, echo 'TYPESAFE_API_KEY=your-key' > .dev.vars, then npm run dev and open http://127.0.0.1:8787.At a glance
- Type
- Demo
- Status
- prototype
- Language
- TypeScript
- Runtime
- Cloudflare Worker with Vite frontend
- Has code
- yes
- Has tests
- yes
Use cases and integrations
How curated lists describe it
Type anything. Get related emojis scored 0–3 with Jev.
hellogumbo/awesome-jev
Frequently asked questions
What does Jev decide?
A 0 to 3 relevance score for emoji groups and candidates against your text.
What is the 'hits' row?
A shortlist built in code from emoji names and keywords, then scored by Jev like the other groups.
Where is the key kept?
Server-side in the Worker, via .dev.vars locally or a wrangler secret in production.
Related projects
Agentic LEGO builder that turns an idea into an LDraw model, using Jev reranking to find suitable parts and example models.
ProjectbetaPython- Jev Search★ 534
Plain-language web search where a decision model like Jev picks query, sources and time range, then ranks Search1API results.
AppproductionTypeScript MCP server over Apple developer docs and WWDC transcripts that uses hybrid search, then Jev relevance scores to rank results.
MCP serverproductionTypeScriptLocal long-term memory for coding agents with an optional Jev reranker that orders recalled memories.
ProjectproductionTypeScriptVolcengine's context database for agents, with a Jev rerank client that scores candidate documents with jev-latest as relevance.
LibraryproductionPython- jev-rerank★ 17
Jev as a calibrated reranker: query plus up to 30 documents in one call, one relevance Noul per document, benchmarked on BEIR.
ProjectbetaPython